Re: [whatwg] Ogg content on the Web

[David Gerard]

On 12/12/2007, Geoffrey Sneddon <[EMAIL PROTECTED]> wrote:
> On 12 Dec 2007, at 14:23, David Gerard wrote:

> > So far we have had zero patent trolls come calling. I wonder why
> > that is.

> Do you have enough money to pay a fine a similar size to what MS got
> last year? If you don't have enough money, they won't sue you. It
> isn't worth their time.


Not to mention "Patent Troll Sues Wikipedia" would be second only to
"Patent Troll Eats Cute Fluffy Kittens" for mediapathy. Mind you, the
people who hate Wikipedia *really hate* Wikipedia, and I'm amazed none
of them have even made noises in this direction, given the
ridiculously broad and vague software and business method patents that
exist in the US.

That said, we do actually go to considerable effort to do the right
thing because it's the right thing - we don't allow patent-encumbered
formats because they would severely reduce the reusability of our
content, and deliberately flouting assumed-valid US patents (as odious
as software patents are) would be unseemly.


- d.

Re: [whatwg] Ogg content on the Web

[David Gerard]

On 12/12/2007, Smylers <[EMAIL PROTECTED]> wrote:

> Not quite.  That's one top-10 source of video that will greatly be
> enabled by browsers supporting Theora.
> Your claim (that it would benefit from the spec saying browsers SHOULD
> support Theora) is only true if there are browsers which would only
> support Theora because of the spec saying that.


Technically this is true :-) But in practice, I can't tell you how
happy we were when we heard Ogg Theora would be in HTML5 (even as a
SHOULD).

Video is important as educational material, and video support in the
MediaWiki software has been a major pain in the backside. Current
support is a kludgy pile of stuff that degrades somewhat gracefully
through a sequence of free-software and not-quite-free-software (VLC
plugin, QuickTime plugin, there's JavaScript, there's a bit of Java,
there's Flash that sorta works in Gnash, etc - I'm not absolutely
clear on the details and I'm sure someone will be along to correct me
shortly, but they're pretty murky details ;-).

A  tag that can be reasonably assumed to support Ogg Theora and
Ogg Vorbis would make our lives and our readers' browsing
significantly happier.


> Some browser creators have made it clear they woudln't support Theora,
> even with a SHOULD.  Other browsers will Theora anyway, because they
> want to, regardless of whether the spec even mentions it -- and the more
> that Wikipedia uses it, the more that browsers are going to want to
> support it simply in order to be Wikipedia-compatible (regardless of
> whether the spec says browsers should be Wikipedia-compatible).


Including, I suspect, Safari - which has a Wikipedia link in the
default bookmark bar - and Nokia - what use is a phone that can't show
you the video on Wikipedia that explains your point precisely when
you're arguing over something in the pub? What sorta rubbishy phone is
that? Tch! Shoulda got an iPhone! *cough*

We're only one site that would significantly benefit from a 
tag that can reasonably be assumed to do Ogg Theora, but we're a
reasonably significant one I think.


- d.

Re: [whatwg] Ogg content on the Web

[Smylers]

David Gerard writes:

> In any case, the point remains: Theora is the only practical option
> for video on Wikimedia sites at present, so that's one top-10 source
> of video that will greatly be enabled for the end user by HTML5 having
> a video tag with Ogg Theora as the default (even as a SHOULD).

Not quite.  That's one top-10 source of video that will greatly be
enabled by browsers supporting Theora.

Your claim (that it would benefit from the spec saying browsers SHOULD
support Theora) is only true if there are browsers which would only
support Theora because of the spec saying that.

Some browser creators have made it clear they woudln't support Theora,
even with a SHOULD.  Other browsers will Theora anyway, because they
want to, regardless of whether the spec even mentions it -- and the more
that Wikipedia uses it, the more that browsers are going to want to
support it simply in order to be Wikipedia-compatible (regardless of
whether the spec says browsers should be Wikipedia-compatible).

Smylers

Re: [whatwg] Ogg content on the Web

[Ralph Giles]

On Wed, Dec 12, 2007 at 07:47:05PM +, David Gerard wrote:

> Dirac is not finished, H.120 has no extant codecs. I may as well call
> "motion PNM" an "unencumbered video format."

Nah. MNG has delta frame compression. Much better than motion PNM. :)

 -r

Re: [whatwg] Ogg content on the Web

[David Gerard]

On 12/12/2007, Geoffrey Sneddon <[EMAIL PROTECTED]> wrote:
> On 12 Dec 2007, at 17:44, David Gerard wrote:
> > On 12/12/2007, Geoffrey Sneddon <[EMAIL PROTECTED]> wrote:
> >> On 12 Dec 2007, at 14:23, David Gerard wrote:

> >>> FWIW, Wikipedia and Wikimedia Commons only allow unencumbered
> >>> formats
> >>> on the site. Video MUST be Ogg Theora. Compressed audio better be
> >>> Ogg.

> >> Why must video just one of many unencumbered formats?

> > Er, what are the others?

> Technically speaking, Theora is actually unencumbered (it just has a
> RF license covering the patents from On2). Dirac is in a similar
> situation.
> Apart from those two, the others I can think of are those that are in
> excess of twenty years old (and therefore their patents have expired),
> such as H.260.


Dirac is not finished, H.120 has no extant codecs. I may as well call
"motion PNM" an "unencumbered video format."

In any case, the point remains: Theora is the only practical option
for video on Wikimedia sites at present, so that's one top-10 source
of video that will greatly be enabled for the end user by HTML5 having
a video tag with Ogg Theora as the default (even as a SHOULD). Claims
that there are no sources of content are simply factually incorrect.


- d.

Re: [whatwg] Ogg content on the Web

[Geoffrey Sneddon]

On 12 Dec 2007, at 19:30, Maik Merten wrote:


Geoffrey Sneddon schrieb:

Apart from those two, the others I can think of are those that are in
excess of twenty years old (and therefore their patents have  
expired),

such as H.260.


I couldn't find anything insightful about "H.260". Sure you don't mean
H.120, which is a 1982 video codec I couldn't find a current
implementation of?


Yeah. I always miscall it H.260 (as it is the precursor to H.261).


H.261, OTOH, is a 1990 standard and thus still a bit away from getting
absolutely free.


Though, by the time we reach LC, it may not be.


--
Geoffrey Sneddon

Re: [whatwg] Ogg content on the Web

[Maik Merten]

Geoffrey Sneddon schrieb:
> Apart from those two, the others I can think of are those that are in
> excess of twenty years old (and therefore their patents have expired),
> such as H.260.

I couldn't find anything insightful about "H.260". Sure you don't mean
H.120, which is a 1982 video codec I couldn't find a current
implementation of?

(Which would explain why Wikipedia is not using it even if it
doesn't happen to hopelessly behind in basically every aspect imaginable)

H.261, OTOH, is a 1990 standard and thus still a bit away from getting
absolutely free.

Maik

Re: [whatwg] Ogg content on the Web

[Geoffrey Sneddon]

On 12 Dec 2007, at 17:44, David Gerard wrote:


On 12/12/2007, Geoffrey Sneddon <[EMAIL PROTECTED]> wrote:

On 12 Dec 2007, at 14:23, David Gerard wrote:


FWIW, Wikipedia and Wikimedia Commons only allow unencumbered  
formats
on the site. Video MUST be Ogg Theora. Compressed audio better be  
Ogg.



Why must video just one of many unencumbered formats?



Er, what are the others?


Technically speaking, Theora is actually unencumbered (it just has a  
RF license covering the patents from On2). Dirac is in a similar  
situation.


Apart from those two, the others I can think of are those that are in  
excess of twenty years old (and therefore their patents have expired),  
such as H.260.



--
Geoffrey Sneddon

Re: [whatwg] Ogg content on the Web

[David Gerard]

On 12/12/2007, Geoffrey Sneddon <[EMAIL PROTECTED]> wrote:
> On 12 Dec 2007, at 14:23, David Gerard wrote:

> > FWIW, Wikipedia and Wikimedia Commons only allow unencumbered formats
> > on the site. Video MUST be Ogg Theora. Compressed audio better be Ogg.

> Why must video just one of many unencumbered formats?


Er, what are the others?


- d.

Re: [whatwg] Ogg content on the Web

[ryan]

On Dec 12, 2007, at 6:23 AM, David Gerard wrote:


FWIW, Wikipedia and Wikimedia Commons only allow unencumbered formats
on the site. Video MUST be Ogg Theora. Compressed audio better be Ogg.

wikipedia.org is something like #8 in the world at present, so this is
set to be a significant content repository actually used by people. A
video tag which can be relied upon to support the format in at least
Firefox would be enormously helpful to us and our readers.

So far we have had zero patent trolls come calling. I wonder why  
that is.


Because the Wikimedia Foundation doesn't have much money. Patent  
trolls are opportunistic, not idealistic.


-ryan

Re: [whatwg] Ogg content on the Web

[Geoffrey Sneddon]

On 12 Dec 2007, at 14:23, David Gerard wrote:


FWIW, Wikipedia and Wikimedia Commons only allow unencumbered formats
on the site. Video MUST be Ogg Theora. Compressed audio better be Ogg.


Why must video just one of many unencumbered formats?

So far we have had zero patent trolls come calling. I wonder why  
that is.


Do you have enough money to pay a fine a similar size to what MS got  
last year? If you don't have enough money, they won't sue you. It  
isn't worth their time.



--
Geoffrey Sneddon